Predicting Protein Interaction Sites Using Structural Similarity
Author Information
Author(s): Monji Hiroyuki, Koizumi Satoshi, Ozaki Tomonobu, Ohkawa Takenao
Primary Institution: Kobe University and Osaka University
Hypothesis
Can we predict interaction sites in proteins with unknown functions by using protein-protein interaction networks and 3D protein structures?
Conclusion
The proposed method has shown effectiveness in predicting interaction sites, especially when applied repetitively.
Supporting Evidence
- The method improves prediction accuracy through a repetitive process.
- Success rates increase with the number of repetitions in predictions.
- The method utilizes both PPI networks and 3D structural information.
Takeaway
This study helps scientists find where proteins interact by looking at similar shapes and features in other proteins.
Methodology
The method involves extracting clusters of neighboring proteins based on structural similarity and predicting interaction sites by comparing pockets on the target protein.
Limitations
The method has only been tested on small datasets and may not perform well on larger networks without modification.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website